Stamping die process data processing method and system based on big data analysis

Abstract

The invention discloses a stamping die process data processing method and system based on big data analysis, and relates to the field of die management, and the method comprises the steps: outputting a crack feature vector; extracting standardized ultrasonic detection data of the stamping die; forming a process comprehensive damage detection result of the stamping die; and dynamically adjusting detection parameters of the electromagnetic field sensor and the ultrasonic sensor. The electromagnetic field frequency and intensity, the ultrasonic transmitting power, the pulse width and the beam focusing are dynamically optimized, adaptive adjustment of parameters of each region is achieved, and it is ensured that the optimal detection effect can be obtained under different material conditions. Meanwhile, closed-loop feedback and multi-source fusion ensure that the system is stable and reliable, the risk of false detection and missing detection is reduced, the controllability and practicability of the detection process are improved, and reliable technical support is provided for efficient and intelligent detection of the stamping die.

Description

Technical Field

[0001] This invention relates to the field of mold management, specifically to a method and system for processing stamping mold process data based on big data analysis. Background Technology

[0002] In industrial production, stamping dies, as crucial forming tools, are widely used in numerous fields such as automotive, aerospace, and electronics. The health of stamping dies directly affects product quality and production efficiency. However, due to the immense pressure, friction, and impact forces exerted on dies during the stamping process, they are prone to surface cracks, internal defects, and other damage. If this damage is not detected and repaired in a timely manner, it can lead to die failure, resulting in production interruptions, product scrap, and other problems, causing significant losses to enterprises. Therefore, accurate and timely damage detection of stamping dies is of paramount importance.

[0003] Wear on molds is often difficult to accurately identify using conventional sensors, especially damage at the microscopic level. Existing technologies employ ultrasonic surface inspection, utilizing the principle of ultrasonic wave propagation. Ultrasonic signals emitted by an ultrasonic sensor are compared with reflected waves from microscopic damage on the mold surface to determine the presence of cracks, wear, or fatigue, without direct contact with the mold surface. However, ultrasonic signals, especially high-frequency signals, experience attenuation as they pass through the mold surface and internal materials, affecting detection accuracy, particularly for detecting micro-cracks and other minute damage. For micro-crack detection, the attenuation of the ultrasonic signal manifests as a decrease in signal strength and a change in the reflected waveform. Micro-cracks, acting as interfaces within the medium, can scatter ultrasonic waves, causing some signals to deviate from the propagation direction, thus leading to signal attenuation. Furthermore, ultrasonic waves are reflected when they encounter micro-cracks. Larger cracks produce stronger reflected waves, but if the crack is very small, the reflected signal may be weak and easily masked by noise, making detection difficult.

[0004] As can be seen from the above, existing technologies for detecting microcracks in stamping dies typically rely on a single ultrasonic testing technique. However, in practical applications, the loads and material stress concentrations during the stamping process often generate fine cracks or micro-delaminations, which are difficult to effectively identify using detection techniques. Although these microcracks or delamination defects are not obvious on the surface, they can lead to localized stress concentrations, thus having a significant negative impact on the long-term service life and structural stability of the die.

[0005] Existing technologies rely on ultrasonic reflection wave analysis for preliminary crack identification. However, when it comes to detecting micro-cracks, problems such as signal attenuation, noise interference, and insufficient accuracy exist, making it impossible to accurately detect damage at the microscopic level. Therefore, it is difficult to effectively prevent mold failure caused by minute defects, which in turn affects production efficiency and product quality. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and system for processing stamping die process data based on big data analysis. To achieve the above objectives, this invention utilizes the following technical solution: a method for processing stamping die process data based on big data analysis, comprising:

[0007] The crack signal on the surface of the stamping die is acquired in real time by an electromagnetic field sensor, and the crack feature vector of the potential defect area of ​​the stamping die is output for preliminary calibration.

[0008] Within the identified potential defect area, a high-frequency acoustic pulse is emitted by an ultrasonic sensor to scan the potential defect area, acquire and analyze the echo signal, and extract standardized ultrasonic test data of the stamping die.

[0009] The crack feature vector and standardized ultrasonic test data are fused together using a data fusion model to form a comprehensive process damage detection result for stamping dies.

[0010] Based on the comprehensive damage detection results of the stamping die process, the defect characteristics of the stamping die surface are analyzed and determined. Based on the defect characteristics, the initial adjustment range of the emission parameters of the electromagnetic field sensor is determined. At the same time, the detection parameters of the ultrasonic sensor are adjusted. Based on the interaction and resonance effect of the electromagnetic field and the ultrasonic beam, the emission parameters of the electromagnetic field sensor are adjusted again.

[0011] In addition, a stamping die process data processing system based on big data analysis is also provided, including:

[0012] The electromagnetic sensing module is used to acquire crack signals on the surface of the stamping die in real time from the electromagnetic field sensor, and output crack feature vectors that preliminarily calibrate the potential defect areas of the stamping die.

[0013] The ultrasonic sensing module is used to scan the potential defect area by emitting high-frequency sound wave pulses through an ultrasonic sensor, acquire and analyze the echo signal, and extract standardized ultrasonic test data of the stamping die.

[0014] The multi-source fusion module is used to fuse crack feature vectors and standardized ultrasonic test data through a data fusion model to form a comprehensive process damage detection result for stamping dies.

[0015] The dynamic adjustment module is used to analyze and determine the defect characteristics of the stamping die surface based on the comprehensive damage detection results of the stamping die process. Based on the defect characteristics, it determines the initial adjustment range of the emission parameters of the electromagnetic field sensor, adjusts the detection parameters of the ultrasonic sensor, and further adjusts the emission parameters of the electromagnetic field sensor based on the interaction and resonance effect of the electromagnetic field and the ultrasonic beam.

[0016] Compared with the prior art, the embodiments of the present invention have at least the following beneficial effects:

[0017] (1) This invention provides a stamping die process data processing method based on big data analysis. It uses an electromagnetic field sensor to collect surface crack signals of the stamping die in real time and outputs crack feature vectors for preliminary calibration. Combined with high-frequency echo scanning from an ultrasonic sensor and extraction of standardized ultrasonic detection data, it achieves precise location of potential defects in the stamping die. After multi-source data fusion, it can obtain comprehensive process damage detection results including defect center location, type classification, depth estimation, reflected energy distribution, and crack propagation direction, thereby significantly improving the accuracy and reliability of die defect identification.

[0018] (2) This invention maximizes the energy density of local defect areas through the interactive coupling and resonance effect of electromagnetic fields and ultrasonic waves, thereby enhancing the amplitude of echo signals from microcracks and tiny defects. At the same time, special adjustment strategies are adopted in low-conductivity material areas, such as optimizing the resonance frequency band, adjusting the ultrasonic emission power and pulse width, to overcome the limitations of insufficient electromagnetic enhancement and improve the detection capability of complex materials and tiny defects.

[0019] (3) This invention analyzes the comprehensive damage detection results of the process and combines the material thickness, conductivity, dielectric constant, magnetic permeability, and defect depth, size, and resolution requirements of each region of the mold to dynamically adjust the detection parameters of the electromagnetic field sensor and the ultrasonic sensor. A multi-objective optimization algorithm is used to calculate the optimal combination of detection parameters for each region, so as to achieve adaptive adjustment of electromagnetic field frequency, intensity, ultrasonic emission power, pulse width, and beam focusing, so that the detection system can maintain the best detection effect under different materials and different defect conditions.

[0020] (4) This invention achieves real-time calibration and optimization of stamping die defect detection by multi-source data fusion and closed-loop feedback adjustment, combined with sensor accuracy, signal-to-noise ratio and historical detection data. The system can set differentiated parameters for different materials and regions to ensure stable and reliable detection results in both high- and low-conductivity materials, reduce false detection and false negative rates, and improve the controllability and practicality of the entire detection process.

[0021] Of course, any product implementing this invention does not necessarily need to achieve all of the above advantages at the same time. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0023] Figure 2 This is a schematic diagram of the system modules of the present invention.

[0024] Figure 3This is a schematic diagram of the logic flow of the present invention.

[0025] Figure 4 This is a flowchart illustrating the dynamic adjustment logic involved in this invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Please see Figure 1 As shown, this embodiment of the invention provides a method for processing stamping die process data based on big data analysis, including:

[0028] like Figure 3 The diagram shown is a logic flowchart involved in the embodiment of the present invention. The diagram illustrates the complete process from electromagnetic field sensor crack signal acquisition, ultrasonic scanning, standardized ultrasonic detection data extraction, multi-source fusion to form a comprehensive damage result, and then to dynamic adjustment of sensor detection parameters and optimization of detection strategy, highlighting the correlation and sequence between each step.

[0029] The crack signal on the surface of the stamping die is acquired in real time by an electromagnetic field sensor, and the crack feature vector of the potential defect area of ​​the stamping die is output for preliminary calibration.

[0030] Electromagnetic field sensors utilize eddy current effects and impedance changes to detect surface defects in stamping dies. When the sensor approaches the die surface, it generates an alternating magnetic field, inducing eddy currents. If defects such as cracks exist on the die surface, the eddy current distribution changes, causing a shift in the sensor's impedance. By detecting these impedance changes, information such as crack location, size, and propagation trend can be output in real time.

[0031] Ultrasonic sensors scan the target area by emitting high-frequency sound pulses. When the ultrasound propagates inside the mold, it encounters cracks or defects, resulting in reflection, refraction, and scattering. By analyzing the propagation time, intensity, and shape of the echo signal, the type, depth, and location of the defect can be inferred. In the calibration area of ​​the electromagnetic field sensor, after receiving the echo signal, the ultrasonic sensor uses signal processing algorithms to extract key information. The depth of the crack or defect is calculated based on the propagation time of the echo signal. The waveform of the reflected signal can determine the morphology of the crack, whether it is transverse or longitudinal, or a defect such as delamination or debonding.

[0032] The electromagnetic field sensor collects crack signals on the surface of the stamping die in real time and outputs crack feature vectors. The specific process is as follows:

[0033] An alternating current is applied to the surface of the stamping die by the excitation coil inside the electromagnetic field sensor, forming an alternating electromagnetic field that induces eddy currents in the conductive die.

[0034] When there are structural defects on the surface of the stamping die, the flow path of the eddy changes, the local current density decreases, and the electromagnetic field distribution in the defect area is disturbed.

[0035] The induction coil of the electromagnetic field sensor receives the electromagnetic response signal after the disturbance, and the electromagnetic response signal changes in the amplitude, phase and impedance parameters of the induced voltage.

[0036] For example, a stamping die is inspected after 1000 stamping operations. The die is made of high-strength steel, the electromagnetic field sensor operates at a frequency of 10kHz and a current of 5A, and the detected current density in a certain area on the die surface ranges from 10A / m. 2 Reduced to 6A / m 2 The voltage amplitude decreased from 0.8V to 0.5V, the impedance increased from 10Ω to 13Ω, and the phase shift was 15°.

[0037] The detection module performs bandpass filtering on the acquired electromagnetic response signal, and then uses fast Fourier transform to perform spectrum analysis on the electromagnetic response signal to extract the amplitude and phase characteristics of the electromagnetic response signal in different frequency bands.

[0038] It should be noted that phase characteristics refer to the phase information of a signal during spectral analysis of an electromagnetic response signal. Specifically, phase refers to the relative positional difference between a specific point in the periodic waveform of a signal and a reference point. In electromagnetic wave and ultrasonic testing, phase characteristics can reflect the signal's propagation process and reflection properties. In electromagnetic testing, especially eddy current testing or other electromagnetic field-based testing methods, the propagation, reflection, and diffraction of electromagnetic waves all cause phase changes. These phase changes reflect characteristics such as the material's electromagnetic properties, the influence of structural defects, and the material's conductivity.

[0039] Based on the complex impedance model, impedance spectrum fitting is performed on the electromagnetic response signal to extract impedance shift features related to crack depth and morphology.

[0040] Impedance offset characteristics refer to the characteristic changes related to material defects obtained by fitting the electromagnetic response signal through a complex impedance model. Specifically, impedance offset refers to the change in the electromagnetic impedance of a material when structural defects are present, reflecting the influence of the defects on the propagation of electromagnetic fields.

[0041] The impedance offset features are normalized and encoded to generate a crack feature vector. The crack feature vector includes the crack existence probability, crack length impedance offset features, crack depth estimate, crack directionality index, and the relative coordinates of the crack on the stamping die surface.

[0042] It should be noted that the real-time acquisition of crack signals on the surface of the stamping die by the electromagnetic field sensor is based on the principles of electromagnetic induction and eddy current detection. Specifically, the excitation coil inside the sensor is supplied with alternating current by the control module, generating an alternating electromagnetic field on the metal surface of the die. According to the law of electromagnetic induction (Faraday's law), this alternating magnetic field induces eddy currents in the conductive die. If there are structural defects such as cracks, inclusions, or pores on or near the surface of the die, the flow path of the eddy currents will be blocked or deflected, and the local current density will decrease, thus forming a magnetic field disturbance in the defect area. This disturbance will cause amplitude attenuation, phase lag, and impedance changes in the electromagnetic signal received by the induction coil or Hall effect sensor.

[0043] The response signal acquired by the electromagnetic field sensor includes time-domain voltage signal and frequency-domain impedance characteristics. The original signal is bandpass filtered to remove environmental electromagnetic interference and power frequency noise. Subsequently, a Fast Fourier Transform (FFT) is used to obtain the amplitude and phase distribution of the electromagnetic response at different frequency bands. To accurately characterize the changes in electromagnetic characteristics caused by defects, the detection module further performs complex impedance spectral analysis on the signal. By fitting the complex impedance model Z(f) = R(f) + jX(f), where R is the resistance component, X is the reactance component, f is the frequency variable representing the electromagnetic signal frequency, and j is the imaginary unit representing the imaginary part of the complex number, impedance shift characteristics related to crack depth and morphology are extracted. Complex impedance refers to the impedance of a circuit or material to alternating current, which includes a real part (resistance) and an imaginary part (reactance). When cracks exist on the mold surface, the local electromagnetic response affects the impedance characteristics of the material.

[0044] After signal feature extraction, a multi-parameter fusion algorithm is used to normalize and encode various features, forming a standardized crack feature vector. The crack feature vector includes the crack presence probability, estimated crack length and depth, crack directionality index, and the relative coordinates of the crack on the mold surface. Specific methods for obtaining this vector include:

[0045] The electromagnetic response signal is fitted using a complex impedance model to obtain the impedance offset value (the difference between the actual impedance and the defect-free reference impedance) at each frequency point. Each frequency point includes the amplitude offset, phase offset, and slope of the impedance spectrum curve. After eliminating differences in the dimensions and magnitudes of different features and performing normalization processing, the normalized features are converted into a unified vector representation, forming a standardized crack feature vector.

[0046] For spatial positioning, two-dimensional or three-dimensional positioning calculations are performed by combining the phase difference of the magnetic fields between sensor arrays. Finally, the crack feature vector output by the electromagnetic field sensor is used as input to provide initial calibration and auxiliary reference for subsequent ultrasonic scanning and quantitative defect identification, realizing a complete detection chain from surface crack identification to structural damage assessment.

[0047] Within the identified potential defect area, a high-frequency acoustic pulse is emitted by an ultrasonic sensor to scan the potential defect area, acquire and analyze the echo signal, and extract standardized ultrasonic test data of the stamping die.

[0048] The standardized ultrasonic test data of the stamping die is extracted, and the specific process is as follows:

[0049] The control module applies an electrical pulse signal to the piezoelectric transducer in the ultrasonic sensor. The piezoelectric transducer generates high-frequency mechanical vibration based on the piezoelectric effect and emits longitudinal or transverse acoustic pulses into the stamping die. The emission frequency is in the range of 1 to 10 MHz. In practical applications, it can be adaptively adjusted according to the characteristics of the die material and the size of the target defect.

[0050] When sound waves propagate in the mold material, they encounter material discontinuities such as cracks, pores, or inclusions. Due to the abrupt change in acoustic impedance at the interface, some sound waves are reflected back to the sensor, forming an echo signal. The reflection coefficient, phase, and energy attenuation characteristics of different types of defects vary.

[0051] Ultrasonic sensors utilize the piezoelectric inverse effect to convert the received echo mechanical vibration into an electrical signal, and record the echo arrival time, amplitude, and waveform characteristics to obtain the echo propagation time, echo amplitude, and waveform distortion. The echo propagation time is used to calculate the defect depth, and the specific expression is as follows: ,in Let v be the defect depth, v be the speed of sound, and t be the echo propagation time.

[0052] Echo amplitude is used to estimate the area of ​​the defect's reflecting surface, while waveform distortion is used to identify the defect type.

[0053] The detection system uses short-time Fourier transform to extract the time-frequency characteristics of the echo from the received signal, and reconstructs the spatial distribution model of the defect through an inversion algorithm. Combined with the potential defect area calibrated by the electromagnetic field sensor, the detection results are matched and calibrated, and standardized ultrasonic detection data is output. The standardized ultrasonic detection data includes the defect center location, defect type classification, defect depth estimation, and reflection energy distribution.

[0054] It should be noted that the ultrasonic sensor in this embodiment of the invention is used for defect detection in stamping dies. Its working principle is based on the propagation characteristics of sound waves in materials and the echo reflection effect caused by the difference in acoustic impedance. Specifically, the control module applies an electrical pulse signal to the piezoelectric transducer element inside the ultrasonic sensor, causing the piezoelectric transducer element to generate high-frequency mechanical vibration based on the piezoelectric effect. This mechanical vibration is then converted into longitudinal or transverse wave sound wave pulses and emitted into the stamping die. The sound waves travel along a certain propagation path in the die material. When they encounter cracks, pores, inclusions, or other material discontinuities, due to the difference in acoustic impedance on both sides of the interface, some of the sound wave energy is reflected at the interface to form an echo signal, while the remaining sound waves continue to propagate. Different types of defects have different reflection coefficients, phase characteristics, and energy attenuation, which can be used to distinguish defect types and characterize defect geometry.

[0055] It should also be noted that the working principle of ultrasonic sensors in stamping die defect detection is based on acoustic propagation characteristics and acoustic impedance reflection mechanism. When the piezoelectric transducer is excited by an electrical pulse, the internal lattice structure of the piezoelectric material deforms under the action of the electric field (i.e., the piezoelectric effect), thereby generating high-frequency mechanical vibration. This vibration is transmitted to the die surface through the coupling medium on the transducer surface, forming ultrasonic pulses in the form of longitudinal or transverse waves.

[0056] An echo signal is a signal formed when an ultrasonic wave encounters an internal defect (such as a crack, delamination, or hole) in a mold and is reflected. By analyzing the amplitude, arrival time, and waveform distortion of the echo signal, the location, size, and type of the defect can be estimated.

[0057] Inversion algorithms refer to the deduction of unknown system structures or parameters from known observation signals. In this embodiment of the invention, the spatial distribution model of defects inside the mold is reconstructed using ultrasonic echo signals, that is, the three-dimensional location and shape of the defects inside the mold are determined. The defect spatial distribution model refers to the representation of the specific location, size, and shape of defects inside the mold using a three-dimensional model or data structure.

[0058] After the echo signal is received by the ultrasonic sensor, the piezoelectric transducer converts the mechanical vibration into an electrical signal through the inverse piezoelectric effect, and records the echo arrival time, amplitude, and waveform characteristics. The echo propagation time is used to calculate the defect depth. The echo amplitude is used to assess the size of the defect reflecting surface and the intensity of reflected energy, while the echo waveform distortion is used to identify the defect morphology and type.

[0059] To extract multidimensional feature information from the echo, the detection module performs short-time Fourier transform or wavelet transform on the received signal to obtain the amplitude, phase, and energy distribution characteristics of the echo signal in the time and frequency domains. Subsequently, an inversion algorithm reconstructs the spatial distribution model of the defect, obtaining the defect's location, depth, and geometry within the mold. To further improve the accuracy of the detection results, the ultrasonic detection results are matched and calibrated with the potential defect area calibrated by the electromagnetic field sensor, thereby outputting standardized ultrasonic detection data, including the defect center location, defect type classification, defect depth estimation, and reflected energy distribution, providing accurate data for stamping die defect identification and damage assessment.

[0060] Continuing the previous example, the ultrasonic sensor frequency was set to 5MHz, and the emitted pulse was a longitudinal wave pulse. The sound wave propagation speed in steel was approximately 590 m / s. The echo amplitude was 0.35 in the cracked area and 0.8 in the crack-free area. The echo propagation time was 0.085 μs. The echo signal showed amplitude attenuation and waveform distortion, indicating that the sound wave propagation path was deflected due to the crack. Through waveform distortion analysis, the detection module determined that the crack was a surface crack.

[0061] The crack feature vector and standardized ultrasonic test data are fused together using a data fusion model to form a comprehensive process damage detection result for stamping dies.

[0062] The crack feature vector from the electromagnetic field sensor and the standardized ultrasonic detection data from the ultrasonic sensor are fused using a big data model to form a comprehensive process damage detection result for the stamping die, specifically including:

[0063] A big data model refers to a data analysis model used to process, analyze, and fuse data from different sensors. In this embodiment of the invention, a data fusion model is adopted, specifically a weighted fusion model.

[0064] The crack feature vector output by the electromagnetic field sensor and the standardized ultrasonic test data output by the ultrasonic sensor are standardized to establish the correspondence between the crack feature vector and the standardized ultrasonic test data.

[0065] The sensor accuracy, signal-to-noise ratio, and historical detection data sets of electromagnetic field sensors and ultrasonic sensors are extracted, processed, and analyzed to obtain the multi-source fusion weights of electromagnetic field sensors and ultrasonic sensors. Geometric transformation is used to align the detection coordinates of electromagnetic field sensors and ultrasonic sensors. Combined with the multi-source fusion weights, data fusion is performed to obtain the comprehensive process damage detection results of stamping dies. The comprehensive process damage detection results of stamping dies include defect center location, type classification, depth estimation, reflected energy distribution, and crack propagation direction.

[0066] In stamping die defect detection, multi-source data fusion aims to jointly analyze crack feature vectors acquired by electromagnetic field sensors and standardized ultrasonic testing data obtained by ultrasonic sensors to improve the accuracy and reliability of defect identification. The specific process includes preprocessing the data acquired by both types of sensors, including noise filtering, outlier removal, and standardization, to ensure the data has a unified temporal and spatial scale for subsequent fusion analysis. The crack feature vectors provided by the electromagnetic field sensors include crack presence probability, crack length impedance offset characteristics, estimated crack depth, crack directionality index, and the relative coordinates of the crack on the stamping die surface. The standardized ultrasonic testing data output by the ultrasonic sensors includes the defect center location, defect type classification, defect depth estimation, and reflected energy distribution.

[0067] In the feature matching stage, a unified feature vector mapping table is established to match crack feature vectors with standardized ultrasonic testing data. A weighted fusion algorithm is used to integrate the matched features, where the weights are adaptively adjusted based on sensor accuracy, signal-to-noise ratio, and historical test results. The fusion method employs probabilistic fusion (such as Bayesian inference) to transform multi-source information into a comprehensive damage index to quantify the probability of defect presence, damage severity, and type.

[0068] By establishing a correspondence between crack feature vectors and standardized ultrasonic testing data, spatial and informational alignment is achieved. Specifically, based on the potential defect areas calibrated by electromagnetic field sensors, ultrasonic echo signals can be mapped to these areas, forming a one-to-one correspondence. In this way, during subsequent fusion, each standardized ultrasonic testing data point can be associated with the spatial location of the electromagnetic crack data, thereby ensuring the accuracy of information fusion.

[0069] In the spatial registration stage, the detection coordinates of the electromagnetic field sensor and the ultrasonic sensor are aligned. Geometric transformation or point cloud registration algorithms are used to ensure the consistency of the two types of data in the spatial position of the mold. The comprehensive evaluation module analyzes the fused data and outputs the comprehensive process damage detection results of the stamping die, including information such as defect center location, type classification, depth estimation, reflected energy distribution, and crack propagation direction.

[0070] The calculation of multi-source fusion weights is based on the accuracy, signal-to-noise ratio (SNR), and historical detection data of the electromagnetic field sensor and ultrasonic sensor. Specifically, the accuracy, SNR, and historical detection results of the electromagnetic field sensor and ultrasonic sensor are first normalized so that their quantized values ​​are all within the range of 0 to 1. The accuracy reflects the reliability of the sensor in detecting target defects, the SNR reflects the clarity of the sensor's output signal, and the historical detection data reflects the sensor's ability to correctly identify defects in previous detections. Then, the three indicators are weighted and summed according to a pre-set adjustment coefficient to form the initial fusion weights of the sensors. Finally, normalization is performed to ensure that the sum of the weights of all sensors is 1. In this embodiment of the invention, the weight formula is expressed as:

[0071] ;

[0072] ;

[0073] in, For multi-source fusion weights of electromagnetic field sensors, The weights for multi-source fusion of ultrasonic sensors. This represents the normalized accuracy value of the electromagnetic field sensor. This represents the normalized signal-to-noise ratio of the electromagnetic field sensor. This is the normalized value of the historical detection data accuracy of the electromagnetic field sensor. This is the normalized value for the accuracy of the ultrasonic sensor. This represents the normalized signal-to-noise ratio of the ultrasonic sensor. This is the normalized value of the historical detection data accuracy of the ultrasonic sensor. The coefficient representing the influence of accuracy index factors. This represents the influence coefficient of the signal-to-noise ratio index. This represents the influence coefficient of historical testing data indicators.

[0074] One example of data is shown in Table 1 below:

[0075] Table 1. Example Data Table

[0076] sensor Precision normalized value Normalized signal-to-noise ratio Historical detection data accuracy Electromagnetic field sensor 0.9 0.8 0.85 ultrasonic sensor 0.95 0.75 0.9

[0077] In this embodiment, for ease of calculation, let... , and The value is 1.

[0078] ;

[0079] ;

[0080] The multi-source fusion weights of the electromagnetic field sensor and the ultrasonic sensor were found to be 0.4951 and 0.5049, respectively.

[0081] After obtaining the weights, the defect feature data output by each sensor are weighted and fused to form a comprehensive process damage detection result for the stamping die. This method quantifies the reliability and signal quality of multi-source information in the fusion result, improving the accuracy of defect identification and providing high-precision data support for subsequent defect localization, type classification, and depth estimation.

[0082] Based on the comprehensive damage detection results of the stamping die process, the defect characteristics of the stamping die surface are analyzed and determined. Based on the defect characteristics, the initial adjustment range of the emission parameters of the electromagnetic field sensor is determined. At the same time, the detection parameters of the ultrasonic sensor are adjusted. Based on the interaction and resonance effect of the electromagnetic field and the ultrasonic beam, the emission parameters of the electromagnetic field sensor are adjusted again.

[0083] like Figure 4 The diagram shown illustrates the dynamic adjustment logic flowchart involved in this invention, providing a clear understanding of the logical relationships and operational sequence between each step. It is used to explain in detail the steps of the entire detection method and their interrelationships.

[0084] Based on the comprehensive damage detection results of the stamping die process, the surface of the stamping die is divided into priority adjustment areas and non-priority adjustment areas, and defect feature vectors are generated for each area.

[0085] The comprehensive process damage detection results include the defect center location, defect type classification number, defect depth, and reflected energy distribution.

[0086] The mold surface is divided into grids or regions, and a region coordinate mapping is established. Based on the comprehensive process damage detection results, the defect characteristics of each region are statistically analyzed, including the number of defects, average depth, and maximum reflected energy.

[0087] When any statistical parameter of a region exceeds the corresponding preset threshold, the region is classified as a priority adjustment region. When all statistical parameters of a region do not exceed the corresponding preset threshold, the region is classified as a non-priority adjustment region.

[0088] Based on the thickness, conductivity, dielectric constant, and magnetic permeability of the mold material in each region, the propagation speed, penetration depth, and attenuation characteristics of electromagnetic waves are analyzed to determine the initial adjustment range of the electromagnetic field emission frequency and intensity. High-frequency electromagnetic waves are used in low-conductivity or thick material regions to reduce attenuation, while low-intensity electromagnetic waves are used in high-conductivity or fragile material regions to prevent thermal effects or damage.

[0089] Based on the defect depth and resolution requirements, the transmission frequency, power, pulse width, and wavelength of the ultrasonic waves are adjusted to adapt to the detection needs of different depths or microcrack areas. Wavelength matching is also performed in conjunction with the electromagnetic field frequency to enhance the coupling effect between electromagnetic waves and ultrasonic waves.

[0090] By adjusting the frequency and intensity of the electromagnetic field, the interaction and resonance effect between the electric field and the ultrasonic beam are achieved, thereby maximizing the energy density of the local defect area.

[0091] A multi-objective optimization algorithm is used to calculate the optimal combination of detection parameters for each region, and the optimized parameters are sent to the electromagnetic field sensor and the ultrasonic sensor, including electromagnetic field frequency, electromagnetic field strength and scanning trajectory, as well as ultrasonic frequency, ultrasonic power, pulse width and array scanning step size.

[0092] It should be noted that the material properties of the mold in different regions may lead to varying responses to electromagnetic waves and ultrasonic waves. Therefore, customized optimization of detection parameters is required for each region. Through multi-objective optimization, the algorithm calculates a set of optimal parameters for each region to ensure the best detection effect within a specific area. The multi-objective optimization algorithm used in this embodiment is the particle swarm optimization algorithm, which is a swarm intelligence-based optimization algorithm suitable for handling multi-objective optimization problems with continuous variables. In practical applications, other multi-objective optimization algorithms can be selected based on the characteristics of the parameters to be optimized.

[0093] Determine the initial adjustment range for the electromagnetic field emission frequency and intensity, specifically including:

[0094] Obtain the material property parameters of the mold material in each region, including the thickness, conductivity, dielectric constant and magnetic permeability of the mold material in each region. Based on the complex dielectric constant, calculate the electromagnetic wave propagation constant and estimate the signal attenuation.

[0095] The complex permittivity describes a material's response to electromagnetic waves, and includes a real part and an imaginary part. The real part represents the material's energy storage capacity, while the imaginary part represents its energy dissipation capacity.

[0096] ;

[0097] in, The complex permittivity, Let be the real part of the complex permittivity. This represents the imaginary part of the complex permittivity.

[0098] ;

[0099] in, For electrical conductivity, It is the frequency of the electromagnetic wave.

[0100] The expression for the propagation constant of electromagnetic waves in a material is:

[0101] ;

[0102] in, The attenuation constant is This is the phase constant.

[0103] The relationship between the propagation constant and the complex permittivity and complex permeability is as follows:

[0104] ;

[0105] in, Angular frequency, . It has negative permeability. It is the constant of pi.

[0106] The decay constant is calculated using the complex permittivity and complex permeability:

[0107] ;

[0108] The calculation process for signal attenuation is as follows:

[0109] ;

[0110] Where G is the quantization representation of signal attenuation, is the attenuation constant, d is the material thickness, and e is the base of the natural logarithm.

[0111] By conducting experiments or electromagnetic simulations (such as the finite element method FEM and the finite difference time-domain method FDTD) under different material and electromagnetic property conditions, the signal response at different frequencies and intensities is measured.

[0112] A regression model is established based on experimental / simulation results. The material properties of the mold material in each region are mapped to the emission parameter mapping set of the electromagnetic field sensor. Based on signal attenuation, the emission intensity and frequency of the electromagnetic field sensor are determined by inputting the corresponding signal attenuation-electromagnetic field sensor parameter mapping set. Based on the material properties of the mold material in each region, the adjustment range values ​​are obtained by mapping to the emission parameter mapping set of the electromagnetic field sensor. Taking the emission intensity and frequency of the electromagnetic field sensor as the center, the initial adjustment range of the emission intensity and frequency of the electromagnetic field sensor is determined. The signal will attenuate due to conductivity, thickness, and dielectric properties when propagating in the material. By estimating the attenuation, the lower limit of the electromagnetic field emission intensity is determined to ensure that there is still enough energy to generate a measurable echo when the signal reaches the defect area, avoiding missed or false detections due to weak signals.

[0113] It should be noted that a regression model is a statistical method used to establish relationships between variables. In this embodiment of the invention, the regression model is used to establish a mapping relationship between the characteristic parameters of the mold material and the emission parameters of the electromagnetic field sensor. The regression model learns the relationship between the properties of the mold material (such as conductivity, dielectric constant, magnetic permeability, etc.) and the emission parameters of the electromagnetic field sensor (such as frequency, intensity, etc.) through experimental data or simulation results. The emission parameter mapping set of the electromagnetic field sensor refers to the set of results obtained after mapping the material characteristic parameters to the emission parameters of the electromagnetic field sensor after the regression model is established. Specifically, the mapping set includes a set of optimal emission parameters that the electromagnetic field sensor should use under different material characteristics in different mold regions.

[0114] It should also be noted that the various thresholds used in this embodiment of the invention are for judging and controlling the numerical characteristics of parameters in the system. A critical point is defined by preset rules. When the parameter value exceeds or falls below the critical point, the system will trigger corresponding operations or adjustments. In this embodiment of the invention, the threshold settings are based on historical data and are continuously adjusted and optimized through multiple simulations and experiments to accurately reflect the critical relationships of parameters under different operating conditions.

[0115] Material thickness directly affects the propagation path and attenuation of electromagnetic waves inside the mold. Areas with greater thickness will cause the electromagnetic wave energy to attenuate significantly during propagation. Therefore, it is necessary to select a higher transmission frequency or appropriately increase the transmission intensity to ensure that the signal can penetrate to the target depth.

[0116] The electrical conductivity of a material determines the energy loss of electromagnetic waves propagating in the mold. Highly conductive materials increase eddy current losses, thus accelerating the attenuation of electromagnetic waves during penetration. To compensate for this attenuation, the system can increase the emission intensity of the electromagnetic field or select electromagnetic waves with higher frequencies, enabling the energy to effectively reach the defect area.

[0117] The dielectric constant and permeability of a material affect the propagation speed, refraction, and reflection characteristics of electromagnetic waves, thereby altering the phase, wavelength, and incident angle of the electromagnetic waves within the mold. By analyzing the distribution of dielectric constant and permeability, the system can predict the propagation mode and attenuation rate of electromagnetic waves in different material regions and determine a suitable frequency range accordingly. This ensures energy concentration of electromagnetic waves in critical defect areas and enhances sensitivity to microcracks or inclusions.

[0118] Taking into account factors such as thickness, conductivity, dielectric constant, and magnetic permeability, a model is established to correlate material properties with electromagnetic field emission parameters. This model is then used to calculate the initial emission frequency and intensity range for each region. This initial adjustment range provides a reference for subsequent dynamic optimization, enabling optimal coupling between the electromagnetic field and ultrasonic signals, and achieving high-precision detection of defects in regions of different thicknesses and materials.

[0119] Based on the defect depth and resolution requirements, adjust the ultrasonic wave's transmission frequency, power, pulse width, and wavelength, specifically including:

[0120] Based on the comprehensive damage test results, the depth information and microcrack size of each region were obtained.

[0121] The thickness of the mold material in each region is mapped to the emission parameter mapping set of the ultrasonic sensor based on a regression model to obtain the corresponding ultrasonic emission frequency. For deeper or thicker material defect areas, a lower ultrasonic emission frequency is selected to enhance penetration. For shallow or microcracked areas, a higher emission frequency is selected to improve spatial resolution.

[0122] The wavelength of an ultrasonic wave is determined based on its transmission frequency. The wavelength and frequency of an ultrasonic wave have the following relationship:

[0123] ;

[0124] in, This refers to the wavelength of ultrasound. The speed at which ultrasound propagates in a material. This refers to the ultrasonic frequency.

[0125] Based on the depth information of each region, the amplitude and signal-to-noise ratio of the echo are analyzed to obtain the echo signal strength. If the echo signal strength is lower than the set low-intensity threshold, it is determined that the transmission power needs to be increased. If the echo signal strength is higher than the set high-intensity threshold, it should be noted that excessively strong signals can lead to ultrasonic transducer saturation or increased noise, thus requiring a reduction in transmission power. Echo signals from deep defects attenuate more during propagation, requiring increased transmission power to ensure measurable signal strength. Lower power can be used for shallow defects or thin material areas. The acoustic properties of the material (such as acoustic impedance, absorption coefficient, and internal structural uniformity) affect the energy attenuation of sound waves. For material areas with high attenuation, the transmission power needs to be increased to ensure sufficient echo signal amplitude. For areas with low attenuation, the power can be reduced to decrease echo distortion or interference caused by excessively strong signals.

[0126] Based on the microcrack size in each region, the signal bandwidth is controlled by adjusting the pulse width. Short pulse widths are used to improve the resolution of small defects, while long pulse widths are used to enhance the energy concentration of deep defects.

[0127] By adjusting the frequency and intensity of the electromagnetic field, the interaction and resonance effect between the electric field and the ultrasonic beam are achieved, thereby maximizing the energy density in the local defect region. The specific processing conditions are as follows:

[0128] At the defect location on the surface of the stamping die, the electric field generated by the electromagnetic field interacts with the ultrasonic beam, and through electro-acoustic coupling and resonance effect, the acoustic energy and electromagnetic energy in the local area are superimposed.

[0129] Based on the material properties of each region and the transmission frequency of the ultrasonic sensor, the control module calculates the optimal frequency offset and readjusts the electromagnetic wave frequency to make the electromagnetic field at the defect location on the stamping die surface phase-matched with the ultrasonic beam.

[0130] Based on the defect depth, material thickness, and ultrasonic echo amplitude, the electromagnetic field emission intensity is readjusted. The specific adjustment process is as follows:

[0131] The electromagnetic field emission intensity is adjusted again based on the defect depth and material thickness. The specific adjustment relationship is simplified as follows:

[0132] ;

[0133] Where I represents the adjusted electromagnetic field emission intensity. The electromagnetic field emission intensity, This is a quantized representation of signal attenuation.

[0134] It should be noted that, in the embodiments of the present invention, the above expressions are simplified expressions used to clarify the relationship between defect depth and material thickness and electromagnetic field emission intensity. In practical applications, calculations should be performed in accordance with physical and mathematical principles.

[0135] There is a positive correlation between the amplitude of the acoustic echo and the intensity of the electromagnetic field emission. The echo amplitude is caused by the excitation of the ultrasonic beam. If the echo amplitude is lower than a set minimum echo threshold, the emission intensity of the electromagnetic field is increased, enhancing the interaction between the electromagnetic wave and the ultrasonic wave. If the echo amplitude is higher than a set maximum echo threshold, the emission intensity of the electromagnetic field is reduced to avoid over-excitation. The absolute difference between the echo amplitude and the corresponding echo threshold is obtained, and this absolute difference is input into the corresponding absolute difference-electromagnetic field emission intensity mapping set to obtain the corresponding electromagnetic field emission intensity.

[0136] The system acquires ultrasonic echo signals in real time, evaluates the electro-acoustic coupling effect through signal processing, and automatically adjusts the electromagnetic wave frequency based on the threshold difference between the phase deviation and the phase deviation threshold if the phase deviation exceeds the phase deviation threshold.

[0137] Specifically, at the defect location on the stamping die surface, there is an electro-acoustic coupling effect between the electric field generated by the electromagnetic field and the ultrasonic beam. When the electromagnetic field acts on the surface or internal defect areas such as cracks and pores of the die, the local electric field will cause polarization, stress changes, or enhanced vibration of the material's microstructure, thereby modulating the propagation characteristics of the ultrasonic wave in that area, including sound velocity, acoustic impedance, and local vibration modes. This modulation can enhance the energy concentration of the sound wave at the defect location. When the frequency of the electromagnetic field is close to the resonance of the ultrasonic beam frequency or its harmonics, the electromagnetic energy and acoustic energy in the local area are superimposed, the echo signal amplitude is enhanced, and the defect boundary and microcrack features are more easily identified.

[0138] Based on the material properties of each defect region (such as conductivity, dielectric constant, magnetic permeability, and thickness) and the emission frequency of the ultrasonic sensor, the electromagnetic wave frequency is optimized a second time to ensure that the electromagnetic field at the defect location is in phase with the ultrasonic beam, thereby maximizing the electro-acoustic coupling efficiency. Through precise phase matching, the local energy concentration effect is enhanced, resulting in a stronger and clearer echo signal.

[0139] Simultaneously, the electromagnetic field emission intensity is adjusted a second time based on the defect depth, material thickness, and real-time received ultrasonic echo amplitude. For deep defects or high-attenuation material regions, the emission intensity is increased to ensure that electromagnetic energy can effectively act on the defect and enhance the echo signal. For shallow defects or low-attenuation regions, the emission intensity is reduced to avoid saturation or interference caused by excessively strong signals. Through this dynamic adjustment based on material properties and real-time signals, the acoustic and electromagnetic energy in local defect regions is optimally superimposed, maximizing the echo signal amplitude and resolution, and improving detection accuracy and reliability.

[0140] It should be noted that during the surface defect detection of stamping dies, the emission frequencies of the ultrasonic sensor and the electromagnetic field sensor are acquired. Based on material properties, the propagation time of the electromagnetic wave and ultrasonic wave from the sensor to the defect location is estimated, thereby calculating the phase difference between the two at the defect location. To achieve phase matching, the electromagnetic wave frequency needs to be adjusted so that its phase at the defect location matches that of the ultrasonic beam, even if the phase difference is close to zero or an integer multiple of 2π. The control module adjusts the emission frequency of the electromagnetic wave by calculating the required frequency offset, so that it achieves synchronous phase with the ultrasonic beam in the defect area, thereby realizing electro-acoustic coupling and energy superposition. Based on the real-time acquisition of echo signals, feedback correction is performed according to the phase deviation. If the phase deviation exceeds the set phase deviation threshold, the electromagnetic wave frequency is further adjusted until optimal phase matching is achieved, thereby maximizing the energy density of the local defect area and improving the sensitivity and accuracy of detection.

[0141] It also includes compensation for the weakening effect of electromagnetic fields on ultrasonic enhancement in low conductivity materials, specifically including:

[0142] In materials with low conductivity, electromagnetic waves may not effectively modulate ultrasonic wave propagation due to the weak eddy current response, resulting in unsatisfactory signal enhancement from microcracks or defects. By increasing the electromagnetic wave emission frequency and appropriately increasing the emission intensity, higher local electric field strength and micro-polarization effects can be generated in the material, thereby improving the electro-acoustic coupling effect.

[0143] When the conductivity of the material in a certain area is lower than a preset threshold, special adjustments are made based on the required conditions, including:

[0144] The optimal resonant frequency band of electromagnetic waves and ultrasonic waves is extracted. The conductivity difference of the material in this region is obtained by subtracting it from a preset threshold. The adjustment parameters of the ultrasonic sensor are obtained based on the conductivity difference, including the transmission power, pulse width and beam focusing parameters of the ultrasonic sensor.

[0145] The conductivity difference reflects the degree of inadequacy in the electromagnetic response capability of the region; a larger conductivity difference indicates a weaker amplification effect of the electromagnetic field on ultrasonic wave propagation. Through pre-established mapping relationships or empirical models, the conductivity difference is converted into adjustment parameters for the ultrasonic sensor. These parameters include transmission power, pulse width, and beam focusing angle. A larger conductivity difference allows for increased ultrasonic transmission power to enhance the penetration ability and echo intensity of ultrasonic waves in low-conductivity regions, thus ensuring effective reception of the defect signal. Based on the conductivity difference, the pulse width is increased to broaden the signal spectrum coverage, making it easier for ultrasonic waves to couple with local electromagnetic waves and increasing energy density. By changing the beam focusing angle or focal length, ultrasonic energy is concentrated in the defect region to compensate for the weakened electromagnetic amplification caused by low conductivity, ensuring maximum acoustic energy from the local defect.

[0146] Please see Figure 2 As shown, the stamping die process data processing system based on big data analysis includes:

[0147] The electromagnetic sensing module is used to acquire crack signals on the surface of the stamping die in real time using an electromagnetic field sensor, and outputs crack feature vectors. The crack feature vectors are used to initially identify potential defect areas of the stamping die.

[0148] The ultrasonic sensing module is used to scan the potential defect area by emitting high-frequency sound pulses from the ultrasonic sensor within the calibrated potential defect area. By analyzing the echo signal, standardized ultrasonic test data of the stamping die is extracted.

[0149] The multi-source fusion module is used to fuse the crack feature vector of the electromagnetic field sensor with the standardized ultrasonic detection data of the ultrasonic sensor through a big data model to form a comprehensive process damage detection result for the stamping die.

[0150] The dynamic adjustment module is used to analyze and determine the defect characteristics of the stamping die surface based on the comprehensive damage detection results of the stamping die process. Based on the defect characteristics, it determines the initial adjustment range of the emission parameters of the electromagnetic field sensor, adjusts the detection parameters of the ultrasonic sensor, and further adjusts the emission parameters of the electromagnetic field sensor based on the interaction and resonance effect of the electromagnetic field and the ultrasonic beam.

[0151] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0152] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to specific implementation methods. Clearly, many modifications and variations can be made based on the content of this specification. The selection and detailed description of these embodiments in this specification are intended to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. Any modifications or variations that do not deviate from the structure of the invention or exceed the scope defined by the invention should fall within the protection scope of the invention.

Claims

1. A method for processing stamping die process data based on big data analysis, characterized in that, include: The crack signal on the surface of the stamping die is acquired in real time by the electromagnetic field sensor, and the crack feature vector of the potential defect area of ​​the stamping die is output in the preliminary calibration. Within the identified potential defect area, high-frequency sound pulses are emitted by an ultrasonic sensor to scan the potential defect area, acquire and analyze the echo signals, and extract standardized ultrasonic test data of the stamping die. The crack feature vector is fused with standardized ultrasonic test data from multiple sources to form a comprehensive process damage detection result for stamping dies; Based on the comprehensive damage detection results of the stamping die process, the defect characteristics of the stamping die surface are analyzed and determined. Based on the defect characteristics, the initial adjustment range of the emission parameters of the electromagnetic field sensor is determined. At the same time, the detection parameters of the ultrasonic sensor are adjusted. Based on the interaction and resonance effect of the electromagnetic field and the ultrasonic beam, the emission parameters of the electromagnetic field sensor are adjusted again.

2. The stamping die process data processing method based on big data analysis according to claim 1, characterized in that: The specific process of acquiring the crack signal on the surface of the stamping die in real time by the electromagnetic field sensor and outputting the crack feature vector is as follows: An alternating current is applied to the surface of the stamping die by the excitation coil inside the electromagnetic field sensor, forming an alternating electromagnetic field, which induces eddy currents in the conductive die. The induction coil of the electromagnetic field sensor receives the electromagnetic response signal after the disturbance, and the electromagnetic response signal changes in the amplitude, phase and impedance parameters of the induced voltage. The built-in detection module performs bandpass filtering on the acquired electromagnetic response signal, and then uses fast Fourier transform to perform spectrum analysis on the electromagnetic response signal to extract the amplitude and phase characteristics of the electromagnetic response signal in different frequency bands. The detection module is a software component used to process and analyze sensor data; Based on the complex impedance model, impedance spectrum fitting is performed on the electromagnetic response signal to extract impedance shift features related to crack depth and morphology. The impedance offset features are normalized and encoded to generate a crack feature vector, which includes the crack existence probability, crack length impedance offset features, crack depth estimate, crack directionality index, and the relative coordinates of the crack on the surface of the stamping die.

3. The stamping die process data processing method based on big data analysis according to claim 1, characterized in that: The specific process for extracting standardized ultrasonic test data of the stamping die is as follows: The built-in control module applies an electrical pulse signal to the piezoelectric transducer element inside the ultrasonic sensor. The piezoelectric transducer element generates high-frequency mechanical vibration based on the piezoelectric effect and emits longitudinal or transverse acoustic pulses into the stamping die. The control module is a software component used to transmit control signals; Ultrasonic sensors utilize the piezoelectric inverse effect to convert the received echo mechanical vibration into an electrical signal, and record the echo arrival time, amplitude, and waveform characteristics to obtain the echo propagation time, echo amplitude, and waveform distortion. The built-in detection module extracts the time-frequency characteristics of the echo using short-time Fourier transform of the received signal, and reconstructs the spatial distribution model of the defect through an inversion algorithm. Combined with the potential defect area calibrated by the electromagnetic field sensor, the detection results are matched and calibrated, and standardized ultrasonic detection data is output. The standardized ultrasonic detection data includes the defect center location, defect type classification, defect depth estimation, and reflection energy distribution.

4. The stamping die process data processing method based on big data analysis according to claim 1, characterized in that: The process of fusing crack feature vectors and standardized ultrasonic testing data through a data fusion model to form a comprehensive process damage detection result for stamping dies specifically includes: The crack feature vector output by the electromagnetic field sensor and the standardized ultrasonic test data output by the ultrasonic sensor are standardized to establish the correspondence between the crack feature vector and the standardized ultrasonic test data. The sensor accuracy, historical detection data set, and signal-to-noise ratio of the electromagnetic field sensor and ultrasonic sensor are extracted, processed, and analyzed to obtain the multi-source fusion weights of the electromagnetic field sensor and ultrasonic sensor. Geometric transformation is used to align the detection coordinates of the electromagnetic field sensor and ultrasonic sensor. Data fusion is performed in combination with the multi-source fusion weights to obtain the comprehensive process damage detection results of the stamping die. The comprehensive process damage detection results of the stamping die include the defect center location, type classification, depth estimation, reflected energy distribution, and crack propagation direction.

5. The stamping die process data processing method based on big data analysis according to claim 1, characterized in that: The comprehensive process damage detection results of the stamping die specifically include: Based on the comprehensive damage detection results of the stamping die process, the surface of the stamping die is divided into priority adjustment areas and non-priority adjustment areas, and defect feature vectors are generated for each area. Based on the thickness, conductivity, dielectric constant and magnetic permeability of the mold material in each region, the propagation speed, penetration depth and attenuation characteristics of electromagnetic waves are analyzed to determine the preliminary adjustment range of electromagnetic field emission frequency and intensity. Adjust the ultrasonic emission frequency, power, pulse width, and wavelength according to the defect depth and resolution requirements; By adjusting the frequency and intensity of the electromagnetic field, the interaction and resonance effect between the electric field and the ultrasonic beam are achieved, thereby maximizing the energy density of the local defect area. A multi-objective optimization algorithm is used to calculate the optimal combination of detection parameters for each region, and the optimized parameters are then sent to the electromagnetic field sensor and the ultrasonic sensor.

6. The stamping die process data processing method based on big data analysis according to claim 1, characterized in that: The determination of the initial adjustment range of the emission parameters of the electromagnetic field sensor specifically includes: Obtain the material property parameters of the mold material in each region, including the thickness, conductivity, dielectric constant and magnetic permeability of the mold material in each region. Based on the complex dielectric constant, calculate the electromagnetic wave propagation constant and estimate the signal attenuation. A regression model is established to map the material properties of the mold materials in each region to the emission parameter mapping set of the electromagnetic field sensor. Based on signal attenuation, the emission intensity and emission frequency of the electromagnetic field sensor are determined. Based on the material properties of the mold materials in each region, the initial adjustment range of the emission intensity and emission frequency of the electromagnetic field sensor is determined with the emission intensity and emission frequency of the electromagnetic field sensor as the center.

7. The stamping die process data processing method based on big data analysis according to claim 1, characterized in that: The adjustment of the detection parameters of the ultrasonic sensor specifically includes: Based on the comprehensive process damage detection results, the depth information and microcrack size of each region were obtained; The detection parameters for ultrasound include transmission frequency, power, pulse width, and wavelength. The thickness of the mold material in each region is mapped to the emission parameter mapping set of the ultrasonic sensor based on the regression model to obtain the corresponding ultrasonic emission frequency, and the ultrasonic wavelength is determined by combining the emission frequency. Based on the depth information of each region, the amplitude and signal-to-noise ratio of the received echo are analyzed to obtain the echo signal strength. If the echo signal strength is lower than the set low strength threshold, it is determined that the transmission power needs to be increased. If the echo signal strength is higher than the set high strength threshold, it is determined that the transmission power needs to be reduced. Based on the microcrack size in each region, the signal bandwidth is controlled by adjusting the pulse width.

8. The stamping die process data processing method based on big data analysis according to claim 1, characterized in that: The emission parameters of the electromagnetic field sensor are further adjusted based on the interaction and resonance effect between the electromagnetic field and the ultrasonic beam. The specific processing conditions are as follows: At the defect location on the surface of the stamping die, the electric field generated by the electromagnetic field interacts with the ultrasonic beam, and through electro-acoustic coupling and resonance effect, the acoustic energy and electromagnetic energy in the local area are superimposed. Based on the material properties of each region and the transmission frequency of the ultrasonic sensor, the control module calculates the optimal frequency offset and readjusts the electromagnetic wave frequency to make the electromagnetic field at the defect location on the stamping die surface phase-matched with the ultrasonic beam. The electromagnetic field emission intensity was adjusted again based on the defect depth, material thickness, and ultrasonic echo amplitude. The system acquires ultrasonic echo signals in real time, evaluates the electro-acoustic coupling effect through signal processing, and automatically adjusts the electromagnetic wave frequency based on the threshold difference between the phase deviation and the phase deviation threshold if the phase deviation exceeds the phase deviation threshold.

9. The stamping die process data processing method based on big data analysis according to claim 1, characterized in that: It also includes compensation for the weakening effect of electromagnetic fields on ultrasonic enhancement in low conductivity materials, specifically including: When the conductivity of the material in a certain area is lower than a preset threshold, special adjustments are made based on the required conditions, including: The optimal resonant frequency band of electromagnetic waves and ultrasonic waves is extracted. The conductivity difference is obtained by subtracting the conductivity of the material in this region from a preset conductivity threshold. The adjustment parameters of the ultrasonic sensor are obtained based on the conductivity difference, including the ultrasonic sensor's transmission power, pulse width, and beam focusing parameters.

10. A system for processing stamping die process data based on big data analysis as described in any one of claims 1-9, characterized in that: The electromagnetic sensing module is used to acquire crack signals on the surface of the stamping die in real time from the electromagnetic field sensor, and output crack feature vectors that preliminarily calibrate the potential defect areas of the stamping die. The ultrasonic sensing module is used to scan the potential defect area by emitting high-frequency sound pulses through an ultrasonic sensor, acquire and analyze the echo signal, and extract standardized ultrasonic test data of the stamping die. The multi-source fusion module is used to fuse crack feature vectors and standardized ultrasonic test data through a data fusion model to form a comprehensive process damage detection result for stamping dies. The dynamic adjustment module is used to analyze and determine the defect characteristics of the stamping die surface based on the comprehensive damage detection results of the stamping die process. Based on the defect characteristics, it determines the initial adjustment range of the emission parameters of the electromagnetic field sensor, adjusts the detection parameters of the ultrasonic sensor, and further adjusts the emission parameters of the electromagnetic field sensor based on the interaction and resonance effect of the electromagnetic field and the ultrasonic beam.

Images